This study adopts a preventive approach to turnover by developing an anticipated turnover concept and measure. If turnover of nursing staff can be anticipated, could strategies be used to intervene before actual turnover occurs? Turnover of nursing staff poses a costly problem to health care; costly both in terms of financial loss and in terms of maintaining a high quality of client care. Prevention of actual turnover is critical. The specific aims of this project are to: 1) To identify the impact of specified individual and organizational factors on anticipated and actual turnover of nursing staff. 2) To estimate the degree to which anticipated turnover predicts actual turnover, particularly voluntary termination of nursing staff. 3) To profile, the characteristics of leavers versus stayers within hospital nursing staff. 4) To describe the relationships between nursing staff characteristics and the specific individual and organizational variables in the Anticipated Turnover Model under voluntary/involuntary; rural/urban; clinical service and staff position conditions. A descriptive, longitudinal, causal modeling design is used to study anticipated and actual turnover is 15 hospitals representing both rural and urban health care facilities. Approximately 2,000 nursing staff from different positions and clinical services will respond to a series of self-report instruments. The istruments have known moderate to strong estimates for reliability and validity. Each staff member will be followed for 12 months in order to document actual turnover. Descriptive, correlation and regression statistics will be used to estimate relationships among the variables. In addition, discriminate analysis will be used to profile the characteristics of the stayers and leavers from the agencies.